It runs in the family – influenza vaccination and spillover effects

We study the Dutch population-based influenza vaccination program, and its spillovers onto family members of those targeted by the program.

In the Netherlands, all individuals aged 65 years and over (60 since 2008) qualify for free influenza vaccination. They receive a personal invitation letter, collect and obtain their flu shot from the GP at zero cost, and the GP receives a remuneration per vaccinated individual of the target group. Those not targeted by the program do not receive an invitation, have to pay out-of-pocket, and first have to collect the vaccine from the pharmacy while only next get the actual flu shot from their GP.

Using a rich dataset that combines survey data on vaccination take-up with information on family linkages from Dutch register data, the quasi-random variation at age 65 is exploited to analyse program impact on vaccination behaviour of family members of targeted individuals. We distinguish between spillovers onto cohabiting partners and spillovers onto adult children. Less than 4 percent of these children still cohabit with their parents, and around 75 percent live within a 20km radius from their partners.

In order to interpret the magnitude of these spillovers, we compare with the direct policy effect on those targeted by the program. Our results indicate a 10 pp increase in vaccination coverage upon crossing the 65-age threshold. The policy has an equally strong 10 pp impact on vaccination take-up among the partner, but only when the partner is younger than 65. We further identified spillovers onto children when the oldest parent turns 65, but not when the youngest parent turns 65. Interestingly, the child spillovers are negative, and about half the size of the direct effects.

We explore several possible mechanisms underlying these oppositely signed spillovers. Epidemiological externalities, changes in perceived prevention benefits and provider-inducement are unlikely candidates. Instead, vaccination patterns are consistent with an information transfer/spillover explanation: partners seem to learn about the benefits of vaccination uptake while children learn they are not targeted by the program. This is further corroborated by the explicit framing of the invitation letter which underlines that individuals aged 65 and over are at increased risk to fall seriously ill due to influenza, and are therefore eligible for free influenza vaccination.

Our findings stress the importance of public health campaigns to pay attention to the effects of information dissemination on public perceptions and attitudes on voluntary preventive care participation. In a companion paper we show that the influenza program reduces mortality, GP visits and prescription drug use among individuals targeted by the program. This paper suggest substantially larger program-associated health effects, although it ultimately depends on the positive health impacts among partners outweighing the negative impact on children, which we could not analyse with the data in this paper.